Www.research.ed.ac.uk



Comparison of accuracy of free-hand, fluoroscopically guided, and aiming device assisted drillingIan Faux BVMS MRCVS- corresponding authorHospital for Small AnimalsUniversity of EdinburghEaster Bush CampusRoslin, MidlothianEH25 9RGianfaux@hotmail.co.uk+447544878909Jon Hall MA VetMB CertSAS DipECVS SFHEA MRCVSVetEd SpecialistsArborbank HouseQuothquanML126NDUnited Kingdom Tobias Schwarz MA Dr med vet DipECVDI DACVR DVR FRCVSHospital for Small Animals University of Edinburgh Easter Bush Roslin, MidlothianEH25 9RGUnited KingdomDylan Neil Clements BSc BVSc PhD DSAS (Orth) SFHEA DipECVS FRCVSHospital for Small Animals University of Edinburgh Easter Bush Roslin, MidlothianEH25 9RGUnited KingdomJohn Ryan MVB CertSAS DipECVS FHEA MRCVSHospital for Small Animals University of Edinburgh Easter Bush Roslin, MidlothianEH25 9RGUnited KingdomWord Count: 3364Abstract:Background: Drilling accuracy is essential in the correct positioning of implants and avoidance of iatrogenic damage to surrounding tissues. The use of augmented drilling methods has been documented as an approach to improving the accuracy of drilling. The aim of this study was to compare the accuracy of two augmented drilling methods (fluoroscopy, aiming device) to free-hand drilling. Methods: Three experienced specialist surgeons and three veterinary surgeons without primary orthopaedic experience drilled into synthetic bone towards a target using the three different methods at three different angles (0o, 10o and 20o). The duration of drilling was recorded, and the accuracy of drilling measured using photographs before and after drilling.Results: The two augmented methods were more accurate than free-hand drilling in synthetic bone with the aiming device producing the greatest accuracy. Increased angulation of drilling decreased the drilling accuracy. Surgeon experience did not impact on drilling accuracy. Surgeon inexperience and augmented drilling methods both increased the time taken to drill.Conclusion: The use of augmented drilling methods improved the accuracy of drilling and surgeons should consider their use when drilling in anatomical regions where the margin of error is small.Introduction:The accuracy of drilling bone is vital in the correct positioning of implants and the avoidance of iatrogenic damage to surrounding soft tissue structures. Damage to the surrounding soft tissue can result in penetration of the joint space (1), spinal cord (2), or major blood vessels. Human error is inevitably a major factor in the accuracy of drilling, as a surgeon is required to identify landmarks, apply guides and/or aiming devices, estimate and maintain angulation whilst drilling through bone. Several of these tasks have been reported to be a cause of inaccuracy (3, 4). Where accuracy is essential such as drilling across the humeral condyle, sacrum, and vertebral body, the benefits of stereotactic instruments, customized drill guides, and intraoperative imaging to maximise accuracy have been documented (5, 6, 7, 8, 9).Fluoroscopy is commonly used as an intraoperative imaging technique in the field of veterinary orthopaedics. Fluoroscopic images enable the real-time static and dynamic intraoperative identification of anatomical structures, which can increase the certainty as to the position of implants as well as facilitating minimally invasive techniques. However fluoroscopic imaging incurs the expense of additional equipment, requires a learning curve, and exposes personnel and patients to ionizing radiation (10, 11).The aim of this study was to compare two augmented methods of bone drilling to ‘traditional’ free hand (palpation-guided) drilling in a synthetic bone model. Our hypotheses were: 1. The two augmented methods would be superior in their accuracy, but that 2. the augmented methods would increase the time taken to drill a hole when compared to free-hand drilling. Materials and Methods: Ethical approval was granted by the University of Edinburgh Human Ethical Review Committee (HERC 320.19). 40mm x 130mm x 180mm, 20 pounds per cubic foot (PCF) polyurethane synthetic bone blocks (Sawbones, Vashon Island, WA, USA) were used to mimic canine cancellous bone, as these are the same density as canine femoral cancellous bone (12).The blocks were held in multi-angle vices (Stanley UK, Slough, UK) to enable tilting at 0 degrees, 10 degrees, and 20 degrees away from the surgeon, and relative to the horizontal table surface. Points were marked using a template created from radiographic films with perforations to allow consistent marking of all blocks. On the upper and lower surface of the block, 20mm x 30mm rectangles were drawn in pencil. On the upper surface a starting point (Point A) was marked on the short edge closest to the surgeon. Five points were marked on the lower surface, consisting of four points spaced 15mm apart in length and 10mm apart in width, and one point (Point B) directly between two peripheral points on the short edge closest to the surgeon (Figure 1). From Point B a distance was measured using the angle of drilling (0 degrees, 10 degrees, 20 degrees from the vertical plane) and the depth of the block (40mm) to mark a target (Point C) on the lower surface of the block for the surgeon to aim for (accommodating for angulation by assuming the drill was vertically positioned vertically when drilling). For the blocks drilled at 0 degrees, Point B and Point C coincide (Figure 1). A small radiopaque marker pin, 2mm in diameter, was then placed at Point C to facilitate palpation and fluoroscopy. The blocks were then tilted to 0, 10, or 20 degrees determined using a digital inclinometer (Trend Direct, Swansea, UK). Photographs were taken of the lower surface of each block before and after drilling using a Canon 750D (Canon: Canon Europa N.V, The Netherlands) with a 70-200f/4L USM lens set in automatic focus and the photos were stored as .jpeg images of 6000x 3368 pixels.Three diplomats of the European College of Veterinary Surgeons and three qualified veterinary surgeons (with no primary orthopaedic experience) drilled into the blocks starting at point A on the upper surface of the block, using a 2.5mm drill bit (310.250: Synthes GmbH, Oberdorf, Switzerland). All participants were required to wear personal protection equipment (PPE) as per our hospital policy. This included lead gowns, thyroid guards, and a personal film badge. In addition, each participant wore an electronic dosimeter beneath the lead gown to monitor radiation exposure in real time and record the overall radiation dose received by each person. All participants had undergone radiation protection in veterinary diagnostic imaging and therapy training, and all were familiar with the safe standard operating procedures of the equipment.Each surgeon aimed for the target pin (Point C) on the lower surface of the block using three techniques: 1. Fluoroscopically-guided (FG), the C-arm (Ziehm 8000, Ziehm Imaging, Nurnberg, Germany) beam was collimated to the minimum required view of the target before introduction of a universal drill guide (2.5mm: Synthes GmbH, Oberdorf, Switzerland) held by the surgeon with their hand out of the primary beam, which was then standardized by viewing the centre of a universal drill guide with the target pin in the middle (Figure 2). The drill guide was held in position and the pilot hole drilled without further fluoroscopic exposure. 2. Free hand drilling (FH), using a universal drill guide and by palpation of the pin on the lower surface of the block alone. 3. Using a universal aiming device (50000, 1.6-4.5mm: IMEX, Texas, USA) (AD), placing the point of the device on point C (without the target pin to prevent slippage). Each technique was repeated three times at each of the specified drill angles in random sequence. For each drilling session, the duration of time taken to complete the task was recorded, starting when the surgeon was notified to begin arranging their equipment (i.e. C-arm, universal aiming device) and ending when the surgeon returned the drill onto the table. The C-arm was positioned in a vertical axis and the synthetic blocks tilted relative to the table and C-arm in order to adhere to a described clinical practice (13) and minimize radiation backscatter (14).Photographs were taken before and after drilling and then superimposed and the marked points on the lower surface of the block aligned (Figure 3) using Adobe Photoshop (Adobe. Adobe Photoshop CC, 19.1.0 ed: Adobe 2019). To assess accuracy of drilling, the drill exit hole location was compared to the target at Point C using coordinates obtained via ImageJ (Image J 1.51; National Institutes of Health, Bethesda, Maryland, USA). The drill holes were outlined using the oval function tool and the centre of hole identified using the centroid pixel coordinates of this oval, as previously described by Bishop et al (15). The distance in pixels was then translated to distance in millimetres by calibration using the known distance between marker points. A student’s t-test was performed on ten measurements randomly selected and no significant differences was found (P-value= 0.280). Therefore, an average for each image was obtained from ten measurements and then used to calibrate the distance from pixels into millimetres. The direction of deviation of the drill hole from Point C was recorded.A Shapiro-Wilk test was performed to assess the normality of distribution of the distance from Point C to the centre of the drill exit holes. A Generalised Linear Mixed Model was used to evaluate the influence of technique, surgeon experience, and angulation on the distance of the centre of drill hole from Point C. Experience, angle and drilling method were included as fixed effects. The interaction of surgeon with angle and surgeon with method were included as random effects. Experience and surgeon were included as different levels to account for the structure of the dataset. The drilling time was also evaluated using a Generalised Linear Mixed Model using the same fixed effects, random effects, interactions, and levels as above. The direction of deviation was categorized using four quadrants: away from and to the right of the surgeon, away from and to the left of the surgeon, towards and the right of the surgeon, and towards and to the left of the surgeon. These quadrants and the pixel coordinates of each drill point were used to create a scatterplot to visually present the data for each surgeon group at each angle (Figure 4). SPSS software was used to perform all statistical analysis (IBM SPSS Statistics for Windows, version 26, IBM Corp., Armonk, N.Y., USA).Results: A total of 27 holes were drilled by each of the six participants, collectively providing 162 data points for analysis. The median distance of the centre of the drill exit point from the target (‘inaccuracy’) when drilled by an experienced surgeon at 0 degrees using the FG method was 1.15mm (range: 0.34-4.46), when drilling by FH method was 2.41mm (range: 0.77-3.55) and when drilling with the AD method was 0.89mm (range: 0.24-1.51).The drilling method was found to be a significant variable affecting the accuracy of drilling (Table 1). An inexperienced surgeon drilling at 20o, by FH was estimated to have a mean drilling inaccuracy of 2.67mm by the statistical model. FH drilling was less accurate then FG drilling (by 0.82mm, P<0.001) and FH drilling less accurate than AD drilling (by 1.40mm, P<0.001). Drilling at 20 degrees reduced accuracy by 0.56mm (P=0.013). Surgeon experience was not found to affect the accuracy of drilling (P=0.772).Model TermCoefficient (mm)Std. Errort-valueSig.95% Confidence Interval Lower UpperIntercept2.6720.2988.976<0.0012.0843.2600 Degree Angle-0.5640.223-5.5250.013-1.005-0.12310 Degree Angle-0.1810.223-0.8100.419-0.6220.26020 Degree AngleReferenceAiming Device (AD)-1.4000.223-6.271<0.001-1.841-0.959Fluoroscopy (FG)-0.8240.223-3.692<0.001-1.265-0.383Free Hand (FH)ReferenceSurgeon Experienced-0.0590.202-0.2900.772-0.4590.341Surgeon InexperiencedReferenceTable 1: Generalized linear mixed model for factors influencing inaccuracy. Reference (intercept) is the inaccuracy (mm) of an inexperienced surgeon drilling by free hand at 20o.Model TermCoefficient(seconds)Std. Errort-valueSig.95% Confidence Interval Lower UpperIntercept27.7168.2013.3800.00111.51843.9150 Degree Angle13.9265.6042.4850.0142.85724.99510 Degree Angle4.9265.6040.8790.381-6.14315.99520 Degree AngleReferenceAiming Device (AD)10.1115.6041.8040.073-0.95821.180Fluoroscopy (FG)81.8525.60414.6070.00070.78392.921Free Hand (FH)ReferenceSurgeon Experienced-12.8525.301-2.4240.016-23.323-2.380Surgeon InexperiencedReferenceTable 2: Generalized linear mixed model for factors influencing time (seconds) taken to drill the hole. Reference (intercept) is the time taken (seconds) by an inexperienced surgeon drilling by free hand at 20o.Drilling method and experience affected the time taken to drill the hole (Table 2). The median duration of drilling when performed AD was 32.5 seconds (range: 11-136), FG was 105 (range: 48-224), and FH was 26 seconds (range: 13-48). The duration of drilling was increased in the inexperienced surgeon group by approximately 13 seconds (P=0.016). When compared to FH, FG drilling took an additional 82 seconds (P=0.000) and AD took an additional 10 seconds (P=0.073).Discussion:Our results support our hypotheses that the augmented drilling methods, using FG or AD, are superior to FH in their accuracy for both experienced and novice surgeons. Augmented drilling methods increase the time taken to drill a hole when compared to FH.In synthetic bone the use of an AD was associated with the greatest accuracy, followed by FG, with FH drilling the least accurate. The level of inaccuracy measured with these techniques was relatively small (0.9mm to 2.4mm) and thus the clinical significance would be dependent on the anatomical location in which the drilling occurs, the size of the patient, and the impact of implant positioning on patient outcome. Although the AD was most accurate when used in this model, other factors present in a clinical setting could not easily be modelled in this study, such as slippage of the aiming device during or after placement and soft tissue, which may impede accurate identification of landmarks used to facilitate drilling in a safe corridor (16). We suggest that FG drilling may be more advantageous in a clinical setting for these reasons.Increasing angulation was found to reduce drilling accuracy, consistent with a previous report (17), although it should be noted that our model examined accuracy when drilling to a target whilst the previously published model examined accuracy of replicating a drilling angle in a single plane. Drilling became statistically significantly inaccurate at 20o, but the magnitude of the differences was small (0.564 mm), with the difference equating to a 0.9o inaccuracy in angulation across the 40mm block. The clinical importance of this degree of inaccuracy may be negligible except in cases where safety margins are minimal (18). The accuracy of angulation was markedly greater in this study than described previously in a different model (17). The high level of accuracy in this study may have been because of the use of a palpable landmark, and that all drilling was undertaken with the drill position vertically rather than at an angle to a horizontal surface. In clinical scenarios with an identifiable target landmark (e.g. the canine humeral condyle) drilling accuracy may be similarly improved compared to those without (e.g. the atlantoaxial joint). A surgeon’s ability to drill accurately to a specified target was not improved with experience in our model, which is consistent with the results of a previous report in which surgeons were asked to drill at a specified angle (17). With a limited number of replications there was little opportunity for the inexperienced surgeons to develop skill acquisition, and thus our model likely describes their innate ability (talent). Our model did not include factors which may have influenced performance of participants in a clinical scenario, such as the psychological stress concerning the consequences of inaccurate drilling, the requirement to identify anatomical landmarks, and to recognize and correct slipping guides on the bone surface. Inexperienced surgeons took longer to complete each drilling procedure, although the method of drilling had a greater effect. Drilling under FG was slowest as the necessary equipment is cumbersome to position. The median times for fluoroscopy and palpation were 105 seconds and 26 seconds respectively. In a clinical setting, these times may be even longer (as the equipment must be set up and draped) but the magnitude of the time difference is unlikely to be of clinical significance. Reported mean surgical duration for fluoroscopic transcondylar screw placement is 35 minutes (range 20-70) (19), whilst the reported mean duration for conventional transcondylar screw placement is 47 minutes (range 20-185) (20).This, when considered in combination with our collected times, highlights that the additional time required to perform fluoroscopy may be compensated for by greater certainty or confidence in where the drill should be placed in the clinical setting, as there is a consequence to inaccurate drilling which is not observed in the model setting.Limitations in this study included the number of participants and sample sizes, which were restricted because of the ethical considerations of using ionising radiation in an experimental study. Therefore, our failure to demonstrate greater accuracy in experienced surgeons may be a type II error or may simply be taken at face value (i.e. experience does not influence accuracy). The synthetic bone blocks were selected to closely mimic a clinical situation but are not a true reflection of the variable density of cortical-cancellous bone interface, although they provide standardised material consistency and avoid variability inherent in cadaveric specimens. The depth of the synthetic bone block was 40mm, this was selected on the basis of its commercial availability and perceived suitability as an ‘average’ depth of drilling of a humeral condyle or sacrum in an adult medium-sized dog breed. The drilling depth may affect the accuracy of drilling and make the use of augmented drilling methods more or less warranted in other clinical scenarios. Visualisation and palpation of landmarks was, potentially, simpler than would be found in a clinical setting.In conclusion, our study indicates that the use of augmented drilling methods (aiming device, fluoroscopy guided) results in more accurate drilling compared with the free hand method, and greater angulation of drilling decreases accuracy. It is therefore important that the surgeon considers the use of these augmented techniques when drilling in anatomical regions where the margin of error is small.References:Clarke SP, Levey J, Ferguson JF. Peri-operative morbidity associated with medio-lateral screw placement for humeral intra-condylar fissure. BVOA proceedings, Birmingham Hilton. 2012; 31-32. Shales C, Moores A, Kulendra E, Toscano M, Langley-Hobbs S. Stabilization of Sacroiliac Luxation in 40 Cats Using Screws Inserted in Lag Fashion. Veterinary Surgery. 2010; 39 696-700.Brioschi V, Cook J, Arthurs GI. Can a surgeon drill accurately at a specified angle? Veterinary Record Open. 2016; 3: e000172.Hernandez D, Garimella R, Eltorai AEM, Daniels AH. Computer-assisted Orthopaedic Surgery. Orthopaedic Surgery. 2017 May; 9(2).Hamilton-Bennett SE, Oxley B, Behr S. Accuracy of a patient-specific 3D printed drill guide for placement of cervical transpedicular screws. Veterinary Surgery. 2018 Feb; 47(2).Sparrow T, Heller J, Farrell M. In vitro assessment of aiming bias in the frontal plane during orthopaedic drilling procedures. Veterinary Record. 2015 Apr 18; 176(16). Ludwig SC, Kramer DL, Balderston RA, Vaccaro AR, Foley KF, Albert TJ. Placement of pedicle screws in the human cadaveric cervical spine: comparative accuracy of three techniques. Spine. 2000 Jul 1; 25(13). Tomlinson JL, Cook JL, Payne JT, Anderson CC, Johnson JC. Closed Reduction and Lag Screw Fixation of Sacroiliac Luxations and Fractures. Veterinary Surgery. 1999 May 1; 28(3). Fujioka T, Nakata K, Nishida H, et al. A novel patient-specific drill guide template for stabilization of thoracolumbar vertebrae of dogs: cadaveric study and clinical cases. Veterinary Surgery. 2019; 48:336–342. Patel R, Dublin E, Olweny EO, Elsamra SE, Weiss RE. Use of Fluoroscopy and Potential Long-Term Radiation Effects on Cataract Formation. Journal of Endourology. 2017; 31 (9) 825-828.Wong L, Rehm J. Radiation Injury from a Fluoroscopic Procedure. New England Journal of Medicine. 2004; 350 (25) e23.Silbernagel JT, Kennedy SC, Johnson AL, Pijanowski GJ, Ehrhart N, D. S. Validation of canine cancellous and cortical polyurethane foam bone models. Vet Comp Orthop Traumatol. 2002(4):200-4.Tonks CA, Tomlinson JL, Cook JL. Evaluation of closed reduction and screw fixation in lag fashion of sacroiliac facture-luxations. Veterinary Surgery. 2008; 37(7):603-7.Miller DL, Balter S, Schueler BA, Wagner LK, Strauss KJ, Vano E. Clinical radiation mangement for fluoroscopically guided interventional procedures. Radiology. 2010; 257(2):321-32Bishop ES, Hall JL, Handel I, Clements DN, Ryan J. Sequential drilling and drill angulation reduce the accuracy of drill hole start location in a synthetic bone model. Veterinary Record. 2019; 184, 383.Barnes DM, Morris AP, Anderson AA. Defining a safe corridor for transcondylar screw insertion across the canine humeral condyle: a comparison of medial and lateral surgical approaches. Veterinary Surgery. 2014; 43(8):1020-31.Brioschi V, Cook J, Arthurs GI. Can a surgeon drill accurately at a specified angle? Veterinary Record Open 2016; 3: e000172. Watine S, Cabassu JP, Catheland S, Brochier L, Ivanoff S. Computed tomography study of implantation corridors in canine vertebrae. J Small Anim Pract. 2006; 47:651-657.McCarthy J, Woods S, Mosley JR. Long-term outcome following management of canine humeral intercondylar fissue using a medial approach and a cannulated drill system. Veterinary Record. 2019; 1-8.Chase D, Sul R, Solano M,Calvo I, Joslyn S, Farrell M. Short- and long-term out-come after transcondylar screw placement to treathumeral intracondylar fissure in dogs. Veterinary Sur-gery. 2019;48:299–308Chase D, Sul R, Solano M,Calvo I, Joslyn S, Farrell M. Short- and long-term out-come after transcondylar screw placement to treathumeral intracondylar fissure in dogs. Veterinary Sur-gery. 2019;48:299–308Chase D, Sul R, Solano M, Calvo I, Joslyn S, Farrell M. Short- and long-term out-come after transcondylar screw placement to treat humeral intracondylar fissure in dogs. Veterinary Surgery. 2019; 48:299–308. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download